Fault Current Article Draws Manufacturer Response
January 1999’s article, “Don’t be felled by higher fault currents,” presents a real situation of substation upgrade that sometimes doesn’t have a simple solution. However, simply placing a current limiting, overcurrent protective device ahead of existing equipment doesn’t really solve the problem.
The author’s intent in this article is to protect a circuit breaker used in a circuit with a higher short-circuit current available than the interrupting rating of the circuit breaker. The method must protect the circuit breaker, conductors, and circuit components. No known calculation method addresses the many factors you must consider, including the current level at which the supply side device becomes protective, performance of the devices in equipment, and the dynamic behavior of the two overcurrent protective devices working together. A miscalculation in any one of these factors eliminates protection. The proposal outlined for the National Electrical Code (NEC) is to replace testing with calculations. This has not proven to be safe.
There is a rigorous series of tests applied to molded-case circuit breakers (MCCBs) used in series ratings. They appear in Underwriters Laboratories Standards for Safety, ANSI/UL 489 and 67. There are also rigorous tests applied to fused, low-voltage power circuit breakers (LVPCBs). These appear in ANSI/IEEE C37.13 and C37.27. Manufacturers who have products with these ratings know they are very expensive to attain and maintain. If a simple method of calculating a workable design with a current limiting device on the supply side existed, circuit breaker manufacturers would want to drop the series ratings and extensive test program in favor of the simpler method.
A method to “engineer” or calculate series ratings without testing has been explored for years. A common misapplied technique is the up-over-and-down method discussed in Fig. 3, on page 30 of the article. This method may have application where truly passive devices are applied. However, the misunderstandings of it are apparent from the extensive discussion over a proposal presented in the 1999 NEC revision cycle. The proposal deems any means to calculate series ratings by “engineering supervision” suitable if accepted by the AHJ. Any calculation method has limitations where circuit breakers are applied. Here are some of them:
Intermediate levels. Fused circuit breakers and MCCBs with series ratings are tested at levels just below the threshold of current limitation for the supply side device. This test is in addition to the test at the maximum short-circuit level. In the article example, 49kA was available with the upgrade. Threshold of current limitation is usually somewhere above 20 times continuous current rating, which would be 40,000A, for a 2000A supply side device. Even if we knew the system was protected for a 49,000A short circuit, we do not know performance below about 40,000A. Would “engineering supervision” know that?
Equipment. Behavior of devices considered individually (outside each electrical enclosure) are not necessarily representative of series conditions. Higher fault availability will generally produce higher peak current. Greater (magnetic) force is placed on the bus structure and insulating conductors than what the electrical equipment has been tested to accommodate. Additional electrical energy may also be delivered through the conductive path, resulting in electrical connection failure due to thermal stress.
Dynamic behavior. Dynamic behavior begins when contacts of the load-side device separate, not when clearing occurs. It’s true some circuit breakers withstand several cycles of current for coordination. However, depending on the trip device settings, contact opening may occur considerably earlier. Without a detailed understanding of contact, mechanism and trip device behavior, an engineer cannot predict protection. Testing is necessary.
Load-side devices. Under the scheme addressed in the article, the whole system has been upgraded, but the solution is directed at only the switchboard closest to the transformer. Faults are most likely to occur at points closer to the loads. Neither the large current limiting device nor the large circuit breaker is likely to operate in time to protect devices and conductors on their load side. Adding a large current-limiting device will do little or nothing to protect the system down from the switchboard.
Well-meaning engineers applied cascade ratings without appropriate tools and testing. ANSI/IEEE C37.13 paragraph 10.8 recommends against cascade ratings. UL standards and the present NEC provide for series ratings, which are tested. Circuit breaker manufacturers do not know of a method that reliably and safely calculates series protection. Until such a method has been found that can be verified by test, short-circuit testing is the only way to engineer series ratings.
George D. Gregory, P.E.
Square D Co.
Cedar Rapids, Iowa
To begin, I know of no fuse or circuit breaker manufacturer that recommends using any engineering approach for selecting series ratings for modern circuit breakers that exhibit dynamic impedance. The industry seems to be in agreement. However, neither the proposed change to the Code nor the article suggests the use of the up-over-and-down engineering method for these modern circuit breakers. Both address the use of an engineering approach for circuit breakers that are passive (no mechanical action) during the first one-half cycle. Decades of safe field experience have shown when the load side circuit breaker is passive while the line side current limiting device is clearing the circuit, the combination can be safely engineered. The article suggests if the circuit breaker manufacturer cannot be reached to determine its passivity, testing labs are available to do so.
The writer points out the components on the load side of the circuit breaker need to be protected under the higher available short-circuit conditions. Absolutely! The engineer must review the short-circuit current ratings of the downstream components to assure they are not exceeded. In fact, the new last sentence of Sec. 110-10 now emphasizes you must research and investigate the product standards, to assure the components are applied within their short-circuit current ratings.
For example, BU 1-1994, the NEMA Busway Standard, states:
the short-circuit rating of busway, which is not marked for use with current-limiting devices, is established on a test basis of three cycles.
That means, it would be a Code violation for busway with a three-cycle rating to be protected by an overcurrent protective device that had a six-cycle short-time delay. It goes on to say:
Busway may be used on circuits having available short-circuit currents greater than the three-cycle rating of the busway rating when properly coordinated with current-limiting devices.
IEEE Standard 141-1993 (Red Book), Electric Power Distribution for Industrial Plants, states in 184.108.40.206:
It is the design engineer’s responsibility to review the current-limiting fuse let-through current for selection of withstand ratings of equipment that has not been tested for a specific class and maximum size of fuse.
Quite possibly, you can find the best reference to backup the article’s material in IEEE Standard 1015-1997 (Blue Book), Applying Low-Voltage Circuit Breakers Used in Industrial and Commercial Power Systems. Paragraph 220.127.116.11(d) states:
In cases where increases in available short-circuit current necessitate a system upgrade, a second approach, shown in Figure 4-4, may be used for retrofitting existing older systems where a recognized series rating is not available. A line-side current-limiting circuit breaker or fuse, which limits peak current and let-through energy, may be added only if the existing load-side breakers do not exhibit dynamic impedance within the first half cycle. The distribution of short-circuit energy is shifted away from the slower, load-side circuit breaker to the higher speed current-limiting device. The downstream circuit breaker is then subject to no more short-circuit energy than its rating. The manufacturer of the existing circuit breakers must be contacted to verify that they do not exhibit dynamic impedance.
You can find information on the intermediate level tests in Table 18.104.22.168 of UL 489. You can also look at a let-through chart to see where a fuse or circuit breaker starts to become current-limiting.
Finally, the last paragraph of Gregory’s letter insinuates the article is suggesting we bring back the cascaded system. This couldn’t be further from the truth. Cascaded systems were a bad design practice in the 1960s, and should never be utilized.
There is no method that the NEC currently blesses for engineering protection of older systems, which have available fault current in excess of the circuit breakers’ interrupting rating. It’s unfortunate many of these older systems are being torn out and replaced by new costly systems—when technically, they could be safely upgraded through the use of current-limiting overcurrent protective devices. Panel 10 passed a solution three times during the last Code cycle, but was rejected on a minor technicality by the Technical Correlating Committee.
Vincent Saporita, P.E.
Director, Training and Technical Services
St. Louis, Mo.